Product Documentation

Résolution des problèmes d'applications

Dec 05, 2017

Real-time application monitoring

You can troubleshoot applications and sessions by using the idle time metric to identify instances that are idle beyond a specific time limit.

Typical use cases for application-based troubleshooting are in the healthcare sector, where employees share application licenses. There, you must end idle sessions and application instances to purge the XenApp and XenDesktop environment, to reconfigure poorly performing servers, or to maintain and upgrade applications. 

The Application Instances filter page lists all application instances on VDAs of Server and Desktop OS. The associated idle time measurements are displayed for application instances on VDAs of Server OS that have been idle for at least 10 minutes. 

Note: The Application Instances metrics are available on Sites of all license editions.

Use this information to identify the application instances that are idle beyond a specific time period and log off or disconnect them as appropriate. To do this, select Filters > Application Instances and select a pre-saved filter or choose All Application Instances and create your own filter.

localized image

An example of a filter would be as follows. As Filter by criteria, choose Published Name (of the application) and Idle Time. Then, set Idle Time to greater than or equal to a specific time limit and save the filter for reuse. From the filtered list, select the application instances. Select option to send messages or from the Session Control drop-down, choose Logoff or Disconnect to end the instances.

Note: Logging off or disconnecting an application instance logs off or disconnects the current session, thereby ending all application instances that belong to the same session.

You can identify idle sessions from the Sessions filter page using the session state and the session idle time metric. Sort by the Idle Time column or define a filter to identify sessions that are idle beyond a specific time limit. Idle time is listed for sessions on VDAs of Server OS that have been idle for at least 10 minutes. 

localized image

The Idle time is displayed as N/A when the session or application instance

  • has not been idle for more than 10 minutes,
  • is launched on a VDA of Desktop OS, or
  • is launched on a VDA running Version 7.12 or earlier.

Historical application failure monitoring

The Trends -> Application Failures tab displays failures associated with the published applications on the VDAs. 

Application failure trends are available for the last 2 hours, 24 hours, 7 days, and month for Platinum and Enterprise licensed Sites. They are available for the last 2 hours, 24 hours, and 7 days for other license types. The application failures that are logged to the Event Viewer with source "Application Errors" are monitored. Click Export to generate reports in CSV, Excel or PDF formats

The grooming retention settings for application failure monitoring, GroomApplicationErrorsRetentionDays and GroomApplicationFaultsRetentionDays are set to one day by default for both Platinum and non-Platinum licensed Sites. You can change this setting using the PowerShell command:

Set-MonitorConfiguration -<setting name> <value>

localized image

The failures are displayed as Application Faults or Application Errors based on their severity. The Application Faults tab displays failures associated with loss of functionality or data. Application Errors indicate problems that are not immediately relevant; they signify conditions that might cause future problems.

You can filter the failures based on Published Application NameProcess Name or Delivery Group, and Time Period. The table displays the fault or error code and a brief description of the failure. The detailed failure description is displayed as a tooltip.

Note: The Published Application name is displayed as “Unknown” when the corresponding application name cannot be derived. This typically occurs when a launched application fails in a desktop session or when it fails due to an unhandled exception caused by a dependent executable.

By default, only faults of applications hosted on Server OS VDAs are monitored. You can modify the monitoring settings through the Monitoring Group Policies: Enable monitoring of application failures, Enable monitoring of application failures on Desktop OS VDAs, and List of applications excluded from failure monitoring. For more information, see Policies for application failure monitoring in Monitoring policy settings.